Skip to content

DevSecOps & Cloud Security

Security in the delivery path: scanning, SBOMs, signing, admission policy, workload identity and posture review.

The problem

Security reviews happen late and by exception. Container images ship with known vulnerabilities, secrets leak into repositories, cloud posture drifts, and the controls that do exist are not enforced where changes actually flow.

What you receive

  • Security gates integrated into the existing pipelines, tuned to be acted on
  • Signed, attested artifacts and a policy that only admits them
  • A posture baseline and the changes that closed the gaps
  • Documentation that maps controls to the frameworks you report against

Capabilities

  • CI security: dependency, container, IaC and secrets scanning that developers act on
  • SBOM generation and image signing with provenance (Sigstore / cosign)
  • Admission policies and runtime controls on Kubernetes
  • Kubernetes and cloud posture review with prioritised remediation
  • RBAC and workload identity: removing static credentials
  • Supply-chain security for the build and deploy path
  • Secure release controls that fit the delivery workflow

Example engagement

Supply-chain hardening for a regulated fintech's Kubernetes platform

Added scanning and SBOMs to the build path, introduced cosign signing with Kyverno admission enforcement, replaced static cloud keys with workload identity, and produced the control mapping their auditors asked for.

Representative of the work; client details are not published.

Technologies

  • Trivy
  • Grype
  • Syft
  • cosign / Sigstore
  • Kyverno
  • Gatekeeper
  • Falco
  • Defender for Cloud
  • AWS Security Hub
  • Workload Identity

Before you write to us

The questions we get asked first.

Answered the way we would on a call. The consulting FAQ covers how engagements work in general.

Where do security controls belong in the pipeline?
Where they can be enforced rather than reviewed: the scan gate before the push, the software bill of materials and signature attached by the pipeline's own identity, and admission control that refuses an unsigned image. A human review is spent on design, not on catching what a policy could decide.
Will this slow the pipeline down?
A scan, a bill of materials and two signatures add a few minutes to a build; admission verification adds milliseconds. The real cost is the first month, when the exception file is long and the base images need updating. After that the gate is quiet because the images are clean.
What access do you need?
Read access to the pipeline, the registry, the cluster's admission configuration, and the current scan results, and a few hours of the team's time. Nothing is run under your production credentials from outside your own tooling, and write access only arrives with an implementation phase, through your pipeline and your review.
What happens after the engagement?
Everything is written down and handed over: the decisions, the code, the runbooks and the reasons. Your team runs it. Where a standing arrangement makes sense, a retainer or a few architect days a month keeps the context without keeping us on the payroll.